Error: Error executing SQL: org.postgresql.util.PSQLException: ERROR: UNION types integer and character varying cannot be matched Position: 325 about legendt2dm

[lofung]

The sql error file says that it is a syntax error for SELECT * FROM #exposure_cohorts;

but r error file says it is a type matching issue, that two error reports do not seem to make sense when they are put together.

We are using postgreSQL

[lofung]

can we solve it this way but how to we solve it for LegendT2DM script?

https://stackoverflow.com/questions/63997315/union-types-text-and-bigint-cannot-be-matched

[lofung]

assessmentLog.txt

2022-01-13 18:19:06	[Main thread]	INFO	CohortDiagnostics	createConceptCountsTable	Creating internal concept counts table

is the last line before it hits the designated error.

[msuchard]

@lofung -- a colleague recently tried the LEGEND-T2DM study code against an postgresql database, and everything appears to work. My best guess is that your tables do not follow the OMOP CDM 5.3 specification. One of these requirements is that cdm_source cannot be null. You can find the specification here (https://github.com/ohdsi/commondatamodel)

The LEGEND-T2DM package depends on many OHDSI packages. So narrowing now your format issues using LEGEND is probably not the easiest. I'd recommend first running something like DataQualityDashboard and a small characterization study first.